Locating a bad resistor

[engels]

I've got a very "beginner's" question.

I just finished a four-triode preamp and the last thing I did was getting rid of all the hum by replacing one 100K ground reference resistor in the second stage. It took me a while to find the bad component and it was a healthy-looking 5% 1W resistor which measured 100K. I've replaced it with a different 100K and the problem was cured.

How do you supposed to locate a bad component without resoldering everything???

[SY]

Logic and experience.

And the latter suggests that the resistor might not have been bad, but the soldering could have been, assuming the new resistor was put back exactly where the old one was. Just for grins, when you have some time, put the old resistor back in. If the hum comes back, we've got an interesting problem (i.e., one that I've never seen before).

[engels]

I understand you're saying it was just a bad solder joint.

Before taking that resistor away I've resoldered the joints and also measured the resistance touching the leads of neighbor components, not the resistor itself.

What could that possible be? I was suspecious about that resistor becuase it was green, so I've replaced it with the same type all the rest of the resistors were (carbon film I believe).

[SY]

The resistor showed 100k in test? Would you happen to have a picture?

edit: Was the ground end of the replacement soldered to the exact same spot? Are there any transformers or inductors nearby?

[richwalters]

No it’s not a silly question. This can decoy many experienced engineers. In my case it wasn’t a component. Sometime ago, that particular problem kept me up burning the midnight oil. <I was experimenting with a (non sharp cut off) pentode (ECF80) as a cathode follower for audio, unaware it was oscillating in the RF and curious to why the DC in the subsequent stages was rising and falling. Unsuspecting initially I thought it was a slack component or valve holder socket but it was the layout at fault and lack of in/out screening. Moving a finger around the circuit proximity or something metallic is the easiest way of detecting RF self oscillation and look at the volts on a following stage decoupled cathode while movement is done- don’t get electric shock by doing this... By chance I put on the screening can and all was solved.
It just so happens the cathode follower configuration is very prone to oscillate esp using ECC88 triodes or any RF tube, one has to be very careful with locating grid stoppers and decoupling caps on pins. When self oscillation occurs, only a few pF is enough, and hum can be picked up,modulated and then demodulated down line. With more RF small signal tubes getting into the audio domain, problems will increase.
Where 99.99% of old TV's have hit the skip, I still have an old 405 line TV, this will pick up weird near-by spurious signals. Don't discard sharp cutoff tubes, these can often get one out of alot of problems.

richj

[richwalters]

Apologies, "beginners" not silly.

richj